noise while making a phone call: hardware or software?

[Mike Montour]

Michael Shiloh wrote:
> I've been searching for this item but all I can find is bug # 883, which 
> says it's a hardware issue, interference making its way into the GSM 
> audio path one way or another.
> 
> Does anyone recall any software explanation for this?

I've started to play around with the mixer settings, and it seems that 
they can be tuned to reduce both the GSM "buzz" and the echo heard on 
the remote end of the phone call. I don't yet have a set of "working" 
state files - the Neo's mixer is quite complicated, and I don't have 
access to the proper test equipment (tone generators, spectrum 
analyzers, etc). However, I've found that a good first start is to set 
to 0 the mixer controls of all of the unused components. There are some 
that are physically not connected (e.g. Amp Mono), and others that are 
unused in certain ALSA profiles (e.g. "Mono Voice" in the gsmhandset 
profile). I will eventually get around to creating a patch for this, but 
if someone else wants to do it first I won't complain.

There is also a hardware side to this issue - if the hardware was really 
good at isolating the audio path from the radio signal, then there 
wouldn't be an interference problem in the first place. However I don't 
think it's reasonable to expect perfection. I can hear a GSM buzz 
similar to the Neo's when my friend calls me from a Motorola RAZR, and 
in general GSM seems to be much more interference-prone than my current 
CDMA phone.

p.s. I'm using http://wiki.openmoko.org/wiki/Neo1973_audio_subsystem and 
its linked pages as my reference. Another TODO for somebody would be to 
annotate the codec block diagram with the ALSA names of the 
corresponding components.